Purpose:
To compare subfoveal choroidal thickness measurements obtained using two different SDOCT instruments in eyes with dry AMD.
Methods:
Dry AMD patients were enrolled into the COMPLETE Study, a prospective randomized, placebo controlled trial designed to investigate the effect of the systemic complement inhibitor eculizumab (Soliris, Alexion Pharmaceuticals) on disease progression in eyes with drusen in the absence of geographic atrophy (Group 1) or eyes with geographic atrophy (Cohort 2). All patients underwent imaging using the Spectralis (Heidelberg Engineering) and the Cirrus (Carl Zeiss Meditec Inc.) SDOCT instruments. For the Spectralis, the enhanced depth imaging protocol utilized a central B-scan that was averaged 51 times. For the Cirrus instrument, the protocol utilized a central B-scan that was averaged 20 times. Two independent graders assessed the images. If the choroidal boundaries could be identified, then these images were classified as good quality and manual choroidal thickness measurements were performed using the caliper software tool available on both instruments. Statistical analyses were performed using the paired t-test, intraclass correlation coefficients (ICC), and Bland-Altman test.
Results:
A total of 60 eyes of 60 patients were included in the study; 30 eyes with drusen and 30 eyes with GA. Mean (SD) age was 70.7 (7.0) in the drusen group and 79.9 (SD=6.6) in the GA group. All images acquired using the Spectralis instrument were classified as gradeable. For the Cirrus instrument, 2 eyes (7%) in the GA cohort with choroids thicker than 400 were classified as ungradeable, while 18 eyes (60%) in the drusen cohort were classified as ungradeable. In the drusen cohort, eyes gradeable by Cirrus had an average Spectralis choroidal thickness of 207.6 µm (48.8), which was significantly thinner (p<0.001) than the average (SD) thickness of those eyes ungradeable by Cirrus, 322.3 µm (76.3). Subfoveal choroidal thickness of the 42 eyes with images gradeable by both instruments averaged 184.9 µm (72.1) by Spectralis and 181.7 µm (75.3) by Cirrus (p=0.19) and the ICC (0.98) was excellent. Bland-Altman analysis did not identify a clear pattern to inter-instrument differences and produced limits of a agreement of ± 31 µm.
Conclusions:
Both the Cirrus and Spectralis SDOCT instruments can be used to measure choroidal thickness; however, the Spectralis instrument was more dependable when imaging the choroid in eyes with different stages of dry AMD.
Clinical Trial:
http://www.clinicaltrials.gov NCT00935883
Keywords: age-related macular degeneration • choroid • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)